projecting problems

[Morten Hulden]

On Sat, 6 Nov 1999, juhana wrote:

> r.proj works occationally but the depending on how the data
>          has been cut it either gives the error that north must be
>          greater than south or then it projects a small piece of the
>          wanted area leaving the rest blank.
> 
> v.proj doesn't give any errors and everything seem ok until
>          data is looked at. Then it may be noticed that the vectors
>           have been turned around to the right direction but are
>           somehow placed having their south edge starting from
>           where the north limit is supposed to be

> Any suggestions?
> Is it a problem with the  projecting program or could there be a
> simpler answer?

It's a problem with r.proj and v.proj. The programs were not designed
to handle other than cylindrical projections and have to be rewritten to
be able to handle the other types. r.proj needs more extensive changes
than v.proj. You may get around the wraparound problem in v.proj by
avoiding to define a full -180-180 region. Leave one of the egdes just one
snap (one resolution unit) below the full value. 


Morten Hulden